ORD Moment of Science: Significantly Reduced Health Burden from Ambient Air Pollution in the U.S. Under Emission Reductions from 1990 to 2010 10/04/2017 Yuqiang Zhang: ORISE Postdoc, ORD/NERL/CED (zhang.yuqiang@epa.gov) ORD/NERL/CED: Rohit Mathur, Christian Hogrefe, Shawn J. Roselle, Jesse O. Bash, Jonathan E. Pleim, Chuen-Meei Gan, David C. Wong Prof. Jason West (UNC-CH) Prof. Jia Xing (Tsinghua University, China)

“Air Quality Improves as America Grows” Introduction “Air Quality Improves as America Grows” Our Nation's Air: Status and Trends Through 2016 MDA8 O3 decreased by 22% from 1990 to 2016; Annual PM2.5 decreased by 42% from 2000 to 2016.

Ambient air pollution and human health Introduction Ambient air pollution and human health Exposure to ozone (O3) and PM2.5 has been associated with both morbidity (e.g. asthma exacerbation) and mortality (e.g. death from cardiovascular and respiratory diseases as well as lung cancer). Estimated 1.04-1.23 million deaths in adults in 2010 globally attributed to long-term O3 exposure (Malley et al., 2017); Estimated 4.2 million deaths attributable to PM2.5 exposure by Global Burden of Disease(GBD), (GBD 2015 Risk Factors Collaborators, 2016).

Ambient air pollution and human health Introduction Ambient air pollution and human health Recent studies have assessed the global (Forouzanfar et al., 2015; Forouzanfar et al., 2016; Lelieveld et al., 2015; Silva et al., 2016) or national (Fann et al., 2012; Punger and West, 2013) burden of disease attributable to air pollution. Few studies have been focusing on the mortality burden trends Cohen et al. (2017) quantified both PM2.5 and O3 mortality burden at global and regional scale from 1990 to 2016 at 5-yr interval; Wang et al. (2017) quantified the PM2.5 mortality burden from 1990 to 2010 in the Northern Hemisphere using Hemispheric scale CMAQ; Fann et al. (2017) quantified the PM2.5 all-cause mortality burden from 1980 to 2010 using geographic kriging PM2.5 from observation

Motivation Quantify the mortality burden from ambient air pollution, including both O3 and PM2.5 in the continental U.S. from 1990 to 2010, based on long-term air quality model simulations by coupled WRF-CMAQ (Gan et al., 2015, 2016) with self-consistent emission inventory (Xing et al., 2013); Analyze contributions of changes in air pollutant concentration, population, and baseline mortality to the overall trend; Analyze the inter-annual variability in mortality estimates.

Methods Model simulation: Coupled WRF-CMAQ two-way model Horizontal resolution of 36 × 36 km covering the Continental U.S. (CONUS); Comprehensive consistent U.S. emission inventory from 1990 to 2010 developed by Xing et al., 2013; Boundary conditions are obtained from 108 × 108 km WRF-CMAQ hemisphere simulation (Xing et al., 2015, Mathur et al., 2017); Simulation period covering 1990 to 2010; Model evaluations for aerosol and O3 have been reported in previous studies (Gan et al., 2015, 2016; Astitha et al, 2017). Gan et al., 2016

For Baseline mortality rates (Y0): Methods Health Impact Function ∆Mort = y0 x AF x Pop ∆Mort: Health burden for O3 or PM2.5; Y0: baseline mortality rates; AF: attributable fraction = (RR – 1)/RR; Pop: Exposed population, ages > 25 yrs For Baseline mortality rates (Y0): (1) Download the death counts for specific disease from the CDC-Wonder; Respiratory diseases (for O3), Lunge cancer (for PM2.5), chronic obstructive pulmonary disease (COPD, for PM2.5), Stroke and ischemic heart disease (IHD) (2) Aggregate the death counts from county level into grid cell (36km x 36km) for each year from 1990 to 2010; (3) The baseline mortality rates for each year was calculated as the total deaths divided by the population. CDC: Centers for Disease Control and Prevention

For attributable fraction (AF): For O3 For PM2.5 Methods Health Impact Function ∆Mort = y0 x AF x Pop ∆Mort: Health burden for O3 or PM2.5; Y0: baseline mortality rates; AF: attributable fraction = (RR – 1)/RR; Pop: Exposed population, ages > 25 yrs For attributable fraction (AF): For O3 Use Log -linear model: RR = expβ∆X RR = 1.040 (1.1013-1.067) per 10 ppbv O3 increases from Jerrett et al., 2009; β: Concentration response factor (3) ∆X = X2(1990-2010) – X1 (Threshold concentration) For PM2.5 Use Integrated Exposure–Response (IER) model: RR equals to z = ambient PM2.5 concentration; zcf = counterfactual conc. = Uniform(5.8,8.8) Model evaluation for the major air pollutants and meteorology variables are presented in several previous studies. Why NH4 is poor—check the reference Underestimating N—Check Wyat’s paper

U.S. air quality trends from 1990 to 2010 Results U.S. air quality trends from 1990 to 2010 6-month average of daily maximum 1hr O3 Annual average PM2.5 For U.S., the population-weighted average and regional average are decreasing for both O3 and PM2.5, which are attributed to the significant emission reductions for the past 2 decades; Population-weighted average/regional average: always larger than 1 for both O3 and PM2.5, which means more people live in high concentration regions.

O3 mortality burden trends from 1990 to 2010 Results O3 mortality burden trends from 1990 to 2010 1. However, they use uniform 22, 30 ppbv for the east and west backgournd o3 concentration. The background O3 should be higher in both the east and the west, larger than 32 ppbv. So they underestimate the background and then overestimate the O3 health burdens O3 mortality burden has inter-annual variabilities, but increase eventually; The national O3 health burdens has increased by 13%, from 109,000 (95%CI, 3,700-17,500) deaths yr-1 in 1990 to 12,300 deaths yr-1 (95%CI, 4,100-19,800) deaths yr-1 in 2010 ;

O3 mortality burden trends from 1990 to 2010 Results O3 mortality burden trends from 1990 to 2010 1. However, they use uniform 22, 30 ppbv for the east and west backgournd o3 concentration. The background O3 should be higher in both the east and the west, larger than 32 ppbv. So they underestimate the background and then overestimate the O3 health burdens If the O3 concentration had stayed constant at 1990 levels (red line), the O3-related mortality burden would have increased by 55% from 1990 to 10,600 (95%CI, 3,600-17,100) deaths yr-1 in 2010; The O3 mortality burden increases were caused by the increases in the baseline mortality rates and population.

PM2.5 mortality burden trends from 1990 to 2010 Results PM2.5 mortality burden trends from 1990 to 2010 1. However, they use uniform 22, 30 ppbv for the east and west backgournd o3 concentration. The background O3 should be higher in both the east and the west, larger than 32 ppbv. So they underestimate the background and then overestimate the O3 health burdens PM2.5 mortality burden has significantly decreased, with smaller inter-annual variabilities; The national PM2.5 mortality burdens have decreased 53% from 123,700 deaths yr-1 (95%CI, 70,800-178,100) in 1990 to 58,600 deaths yr-1 (95%CI, 24,900-98,500) in 2010.

PM2.5 health burden trends from 1990 to 2010 Results PM2.5 health burden trends from 1990 to 2010 1. However, they use uniform 22, 30 ppbv for the east and west backgournd o3 concentration. The background O3 should be higher in both the east and the west, larger than 32 ppbv. So they underestimate the background and then overestimate the O3 health burdens The PM2.5-related mortality burden in 2010 would have decreased by only 24% (94,400 deaths yr-1 in 2010, 95%CI, 50,300-139,800) compared with that in 1990, if the PM2.5 concentrations had stayed constant over the period 1990-2010 (red line); The PM2.5 mortality burden decreases were attributed to the decrease of PM2.5 concentration and baseline mortality rates.

Conclusion Both O3 and PM2.5 are decreasing in U.S. for the past two decades, especially during the east and west coast; The O3 health burdens are increasing, mainly caused by the increases of the baseline mortality rates and population changes; However, the O3 decreases under the emission reductions for the past 2 decades have significantly slowed down the O3 mortality burdens increase trends (increase by 13% compared with the increases by 55% from 1990 to 2010); The PM2.5 mortality burdens are decreasing, as a combined effect of decreased baseline mortality rates and concentration; The PM2.5 decreases under the emission reductions have increased the PM2.5 health burden decrease trends (decreases by 53% compared with decreases by 24% from 1990 to 2010).

Acknowledgements We would like to thank Neal Fann from OAR/OAQPS/HEID for the useful discussions on the county-level baseline mortality data from the CDC website. Thank Dr. Raquel Silva from ORD/NERL/EMMD/PHCB for the mortality burden analysis.

Thanks for your attention! Questions and Comments?

Extra slides

For exposed population (ages > 25yrs): Methods ∆Mort = y0 x AF x Pop ∆Mort: Health burden for O3 or PM2.5; Y0: baseline mortality rates; AF: attributable fraction = (RR – 1)/RR; Pop: Exposed population, ages > 25 yrs For exposed population (ages > 25yrs): 2000 2010

For baseline mortality rates: Methods ∆Mort = y0 x AF x Pop ∆Mort: Health burden for O3 or PM2.5; Y0: baseline mortality rates; AF: attributable fraction = (RR – 1)/RR; Pop: Exposed population, ages > 25 yrs For baseline mortality rates: 2000 2010

Methods Baseline mortality rates The baseline mortality rates for cause-specific death related with PM2.5, including chronic obstructive pulmonary disease (a), lung cancer (b), ischemic heart disease (c) and stroke (d), and the Respiratory diseases (e) related with O3. The bottom whiskers, bottom border, middle line, top border and the top whiskers of the boxes, indicate the 5th, 25th, 50th, and 75th, 95th percentiles, respectively, across all counties; The red circles are the national-level rate. Baseline mortality rates are shown for 1990-1998 after they are corrected to ensure comparability between ICD9 and ICD10 codes. Model evaluation for the major air pollutants and meteorology variables are presented in several previous studies. Why NH4 is poor—check the reference Underestimating N—Check Wyat’s paper

U.S. emission trends from 1990 to 2010 Xing et al., 2013 Use a consistent framework across years to develop U.S. emissions estimates from 1990 to 2010. Emissions for major air pollutants are greatly decreasing, except for NH3, the increase of which is mainly caused by the activity of livestock and agriculture.

U.S. air quality trends from 1990 to 2010 Results U.S. air quality trends from 1990 to 2010 6-month average of daily maximum 1hr O3 O3 decreases significantly in the east and west, > 0.5 ppbv yr-1; The decreases are slow or insignificant in the middle; PM2.5 decreases in the Northeast, but no significant changes in the west. Annual average PM2.5

Methods Background O3 (1850): ∆Mort = y0 x AF x Pop Model evaluation for the major air pollutants and meteorology variables are presented in several previous studies. Why NH4 is poor—check the reference Underestimating N—Check Wyat’s paper From an ensemble of 14 global chemistry–climate models (ACCMIP; Silva et al., 2013) O3 higher in the eastern and western U.S., large than 30 ppbv; lower in the middle, larger than 28 ppbv

Trends for the absolute contribution of the three factors Results Trends for the absolute contribution of the three factors 1. However, they use uniform 22, 30 ppbv for the east and west backgournd o3 concentration. The background O3 should be higher in both the east and the west, larger than 32 ppbv. So they underestimate the background and then overestimate the O3 health burdens

Compared with previous studies Results Compared with previous studies Comparisons of the U.S. mortality burdens attributed to PM2.5 (a), and O3 (b) in this study, with Cohen et al. (2017), Fann et al. (2012), Punger and West (2013), and Giannadaki et al. (2017). The black dashed line shown in the O3 health burden comparisons is the recalculated O3 health burdens using the pre-industrial O3 concentration. The error bars stand for the 95% CI from the RRs, shown for this study and Cohen et al. (2017).